Article type
Abstract
"Background:
The relative treatment effect is often considered to be portable (or constant) across populations with different baseline risks, to the extent that clinical decision-making and guideline recommendations are based on absolute effects derived from relative effects. We sought to determine in a very large sample of meta-analyses how often the relative treatment effect varied among populations with low, moderate and high baseline risk.
Methods:
We retrieved data from all meta-analyses published in the Cochrane Database of Systematic Reviews (2003-2020) that evaluated a binary outcome, reported 2x2 data for each individual study and included at least 4 studies. We excluded studies with no events. We conducted meta-analyses with odds ratios and relative risks and performed subgroup analyses based on tertiles and quartiles of the baseline risk.
Results:
The analysis included 2,531 systematic reviews (27,692 meta-analyses, 226,975 studies, 25,669,783 patients).The percentages of meta-analyses with statistically significant interaction based on baseline risk quantile ranged 12-25% across various sensitivity analyses. This percentage was higher with the fixed-effect model than random-effects models and increased as the number of studies or range of baseline risk per meta-analysis increased.
Conclusion:
Analysis of a very large sample of studies suggests that relative treatment effects are not portable across populations with varying baseline risks in a substantial number of meta-analyses, perhaps 1 in 5. Guideline developers and decision-makers should be provided with relative and absolute treatment effects that are conditioned on baseline risk or derived from studies with similar baseline risk to their target populations."
The relative treatment effect is often considered to be portable (or constant) across populations with different baseline risks, to the extent that clinical decision-making and guideline recommendations are based on absolute effects derived from relative effects. We sought to determine in a very large sample of meta-analyses how often the relative treatment effect varied among populations with low, moderate and high baseline risk.
Methods:
We retrieved data from all meta-analyses published in the Cochrane Database of Systematic Reviews (2003-2020) that evaluated a binary outcome, reported 2x2 data for each individual study and included at least 4 studies. We excluded studies with no events. We conducted meta-analyses with odds ratios and relative risks and performed subgroup analyses based on tertiles and quartiles of the baseline risk.
Results:
The analysis included 2,531 systematic reviews (27,692 meta-analyses, 226,975 studies, 25,669,783 patients).The percentages of meta-analyses with statistically significant interaction based on baseline risk quantile ranged 12-25% across various sensitivity analyses. This percentage was higher with the fixed-effect model than random-effects models and increased as the number of studies or range of baseline risk per meta-analysis increased.
Conclusion:
Analysis of a very large sample of studies suggests that relative treatment effects are not portable across populations with varying baseline risks in a substantial number of meta-analyses, perhaps 1 in 5. Guideline developers and decision-makers should be provided with relative and absolute treatment effects that are conditioned on baseline risk or derived from studies with similar baseline risk to their target populations."